Fuel cells are gaining the attention of a broad set of mobile systems developers and investors as the best hope for extending battery life beyond the capabilities of today's lithium ion batteries. Nevertheless, the approaches of both start-ups reveals the many complex and difficult hurdles the technology faces.

"Lithium ion is plateauing in its performance, and there is no new battery chemistry on the horizon," said David Dorheim, the president and chief executive of Neah (Bothell, Wash.).

"We've talked to a lot of people in battery design, and there's nothing else with the density required," agreed Jim Balcom, president and chief executive of Polyfuel (Menlo Park, Calif.).

More than two dozen investors packed presentation rooms to hear from both companies at the recent Wireless Ventures conference here (May 1).

"When these companies came around looking for Series A and B funding over the last couple years we saw them as science projects, and we don't invest in science projects. But now it looks like this could be a very big thing," said Stewart Alsop, a well-known venture capitalist and columnist for Fortune magazine.

Nevertheless, for fuel cells to succeed they must be similar to lithium ion in form factor, competitive in price and superior in energy density—three tests no system has passed to date. In addition, fuel cell makers must plow aside and regulatory issues and create an infrastructure to deliver the disposable fuel cartridges they require.

A number of startups are pursuing direct-ethanol fuel cells that pump a mixture of water and methanol across electrodes mated to a permeable membrane, but to date those cells have not shown significantly higher energy density than lithium ion batteries. Polyfuel hopes to change that with a highly efficient membrane it will ship to in July based on research spun out of SRI (Menlo Park) in 1999.

Balcom said fuel cells based on the new membrane could deliver 10-12 hour batteries for notebooks, something beyond the reach of lithium ion. However, the cells would not compete with the two- and four-hour batteries today's notebooks use, he added. The company got a green light in the U.S. last year to ship on airplanes the 25 percent methanol solution its approach requires. However getting international approvals through the United Nations could take more than two years.

Separately Polyfuel is in talks with lighter and pen maker BIC and battery maker Duracell about creating a retail market for $2 to $3 disposable methanol cartridges that would power fuel cells.

Even if Polyfuel can surmount those hurdles, fuel cell makers need to significantly miniaturize today's large prototype membrane-based fuel cells that include a pump, fan and small battery for initializing the pump action. In addition, OEMs likely would have to build systems that incorporate some type of small lithium ion battery to initiate the fuel cell pumping action and power systems during spikes when peak power is needed.

Lithography technique

Neah takes a very different approach to solving the energy density problem. The company plans to use lithography to help drill meshes through slivers of specially grown and treated silicon wafers, then dice and stack the chips into cells that have five to ten times the reactive surface area of membrane-based cells. The result could be fuel cells with two or three times the energy density of lithium ion batteries using designs that can ride on improvements in semiconductor manufacturing technology, said Dorheim.

To succeed, Neah must strike a partnership with a company that grows silicon wafers and perfect a modified semiconductor manufacturing process. It must also help build its own infrastructure of methanol cartridges that meet its unique requirements that include providing hydrogen peroxide as an oxidant and trapping waste and by-products of the reaction in the cartridge.

Neah is developing its first fuel cell prototypes this year. It then plans to engage an OEM systems partner next year and deliver its first vertical market designs sometime in 2005. Sales to mainstream notebook PC markets are not expected until 2006, after it has built up an infrastructure for its retail cartridges.

Neither company expects to get the cells, which are as large as PDAs in some current prototypes, into cellular phones anytime soon. "Mobile phones are a very difficult application for fuel cells. For larger PDAs or smart phones-perhaps," said Dorheim. Balcom said he helped Polyfuel focus on being only a supplier of membranes, rather than try to build whole fuel cells and the infrastructure they require.

"To go the route of developing the whole system is futile. You end up making big systems companies like Toshiba and NEC your competitors," Balcom said.

Balcom said a host of large companies including Casio, Intel, Motorola, NEC, Panasonic, Samsung and Toshiba have announced active fuel cell programs for mobile systems. Toshiba has publicly stated it plans to commercialize fuel cells by late 2004.

Market watcher Allied Business Intelligence (Oyster Bay, N.Y.) estimates as many as 50 million fuel cells for portable systems could ship in 2005, growing to 200 million units in 2008. "It's not a question of if, but when," said Balcom.